Morgan Stanley has raised its 2026 capital expenditure forecast for the largest cloud hyperscalers to $805 billion, up from $765 billion, in the latest sign that the AI infrastructure buildout is accelerating rather than cooling.
The figure aggregates projected spending from Microsoft, Amazon, Google, Meta, and Oracle, the five companies pouring the most capital into the data centers, networking, and accelerator chips that underpin generative AI services. The upward revision lands as each of those companies has used recent earnings calls to signal that demand still outstrips available compute.
The forward number is even more striking. Morgan Stanley analyst Ben Reitz flagged that 2027 spending could hit a $1.1 trillion mark, a level that would match the combined capex of every non-tech company in the S&P 500. That would equal all non tech S&P 500 cap combined, that is insane, Jaeden Schafer said on the podcast, framing the projection as a once-in-a-cycle reordering of where corporate America puts its money.
Key facts
- 01Morgan Stanley raised its 2026 hyperscaler capex forecast to $805 billion, up from $765 billion previously.
- 02The forecast covers combined spending from Microsoft, Amazon, Google, Meta, and Oracle on AI data center buildouts.
- 03Analyst Ben Reitz projects 2027 hyperscaler capex could reach $1.1 trillion, matching all non-tech S&P 500 capex combined.
- 04White House AI czar David Sacks told Benzinga that AI could account for as much as 75% of US GDP growth in coming years.
On AI Chat Daily, Jaeden Schafer argued the trajectory has effectively turned the hyperscalers into a parallel industrial economy. "AI infrastructure is basically becoming as much, you know, capex as the rest of the, you know, entire corporate economy outside of tech," he said. The implication is that AI compute is no longer a line item inside Big Tech budgets — it is approaching the scale of every factory, pipeline, and utility build outside the sector.
“AI infrastructure is basically becoming as much, you know, capex as the rest of the, you know, entire corporate economy outside of tech.”— Jaeden Schafer
The political messaging around those numbers is escalating in parallel. White House AI czar David Sacks, speaking to Benzinga, suggested AI could account for as much as 75% of US GDP growth over the coming years, a claim that would put a single technology cycle at the center of the country's economic story.
Schafer treated the figure with some skepticism about the messenger but not the direction of travel. He noted Sacks is "talking his own book" given his role overseeing AI policy, while also pointing out that recent growth data already lines up with the thesis. Over the past three years, hyperscaler capex tied to AI and data centers has been one of the dominant drivers of US business investment, making outsized GDP contribution less of a stretch than the headline number suggests.
The bull case Sacks is implicitly pitching goes further than the buildout itself. The argument is that once AI systems are deployed across the economy, productivity gains will translate the current capex wave into broad-based output growth — not just revenue for the companies selling chips and cloud capacity.
For investors, Morgan Stanley's revision tightens the link between the AI trade and the broader market. With five companies on track to spend more than $800 billion in a single year, any wobble in expected returns from that capital — whether from slower enterprise adoption, regulatory friction, or chip supply — would ripple far beyond the tech sector. For now, the bank's forecast points the other way: spending plans keep getting bigger, not smaller.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
